The General Theory of Employment, Interest, and Money

 W
(
Y
w
). 
The amount that the community spends on consumption obviously depends (i) partly on the amount 
of its income, (ii) partly on the other objective attendant circumstances, and (iii) partly on the 
subjective needs and the psychological propensities and habits of the individuals composing it and 
the principles on which the income is divided between them (which may suffer modification as 
output is increased). The motives to spending interact and the attempt to classify them runs the 
danger of false division. Nevertheless it will clear our minds to consider them separately under two 
broad heads which we shall call the subjective factors and the objective factors. The subjective 
factors, which we shall consider in more detail in the next chapter, include those psychological 
characteristics of human nature and those social practices and institutions which, though not 
unalterable, are unlikely to undergo a material change over a short period of time except in 
abnormal or revolutionary circumstances. In an historical enquiry or in comparing one social system 
with another of a different type, it is necessary to take account of the manner in which changes in 
the subjective factors may affect the propensity to consume. But, in general, we shall in what 
follows take the subjective factors as given; and we shall assume that the propensity to consume 
depends only on changes in the objective factors. 
II 

---

50
The principal objective factors which influence the propensity to consume appear to be the 
following: 
(1) 
A change in the wage-unit
. Consumption (
C
) is obviously much more a function of (in some 
sense) 
real
income than of money-income. In a given state of technique and tastes and of social 
conditions determining the distribution of income, a man's real income will rise and fall with the 
amount of his command over labour-units, i.e. with the amount of his income measured in wage-
units; though when the aggregate volume of output changes, his real income will (owing to the 
operation of decreasing returns) rise less than in proportion to his income measured in wage-units. 
As a first approximation, therefore, we can reasonably assume that, if the wage-unit changes, the 
expenditure on consumption corresponding to a given level of employment will, like prices, change 
in the same proportion; though in some circumstances we may have to make an allowance for the 
possible reactions on aggregate consumption of the change in the distribution of a given real income 
between entrepreneurs and rentiers resulting from a change in the wage-unit. Apart from this, we 
have already allowed for changes in the wage-unit by defining the propensity to consume in terms 
of income measured in terms of wage-units. 
(2) 
A change in the difference between income and net income
. We have shown above that the 
amount of consumption depends on net income rather than on income, since it is, by definition, his 
net income that a man has primarily in mind when he is deciding his scale of consumption. In a 
given situation there may be a somewhat stable relationship between the two, in the sense that there 
will be a function uniquely relating different levels of income to the corresponding levels of net 
income. If, however, this should not be the case, such part of any change in income as is not 
reflected in net income must be neglected since it will have no effect on consumption; and, 
similarly, a change in net income, not reflected in income, must be allowed for. Save in exceptional 
circumstances, however, I doubt the practical importance of this factor. We will return to a fuller 
discussion of the effect on consumption of the difference between income and net income in the 
fourth section of this chapter. 
(3)